Modern wireless communication systems such as GSM (Global System for mobile communications) and UMTS (Universal Mobile Telecommunications System) are presumed to transfer various types of data over the air interface between the network elements such as a base station and a mobile station (MS). As the general demand for transfer capacity continuously rises due to e.g. new multimedia services coming available, new more efficient techniques have been developed in order to exploit the existing resources to a maximum extent.
A technical report 3GPP 45.902 [1] discloses a concept of flexible layer one, a new physical layer proposed for the GERAN. The ingenuity of the concept relies on the fact that the configuration of the physical layer including e.g. channel coding and interleaving is specified not until the call set-up. Thus, the support of new services can be handled smoothly without having to specify new coding configuration schemes separately in connection with each release.
Development work of the FLO concept has been provided with somewhat strict requirements. FLO should, for example, support multiplexing of parallel data flows on to a basic physical subchannel and provide optimization of spectral efficiency through the support of different interleaving depths, unequal error protection/detection, reduced channel coding rate granularity and support of different (8PSK, GMSK etc) modulations. Moreover, the solution shall be future proof and minimize the overhead introduced by the radio protocol stack.
According to the GERAN Release 5 the MAC sublayer (Layer 2 for FLO) handles the mapping between the logical channels (traffic or control) and the basic physical subchannels introduced in 3GPP TS 45.002 [2].
In UTRAN (UMTS Radio Access Network), the MAC utilizes so-called Transport Channels TrCH for transferring data flows with given QoS's (Quality of Service) over the air interface. As a result, several transport channels, that are configured at call set-up, can be active at the same time and be multiplexed at the physical layer.
Now, by adopting the idea of FLO, aforesaid flexible transport channels can be utilized in GERAN as well. Accordingly, the physical layer of GERAN may offer one or several transport channels to the MAC sublayer. Each of these transport channels can carry one data flow providing a certain Quality of Service (QoS). A number of transport channels can be multiplexed and sent on the same basic physical subchannel at the same time.
The configuration of a transport channel i.e. the number of input bits, channel coding, interleaving etc. is denoted as a Transport Format (TF). Furthermore, a number of different transport formats can be associated to a single transport channel. The configuration of the transport formats is completely controlled by the RAN (Radio Access Network) and signalled to the MS at call set-up. Correct interpretation of the TF is crucial at the receiving end as well as the transport format defines the utilized configuration for decoding of the data. When configuring a transport format, the RAN can, for example, choose between a number of predefined CRC (Cyclic Redundancy Check) lengths and block lengths.
On transport channels, transport blocks (TB) are exchanged between the MAC sublayer and the physical layer on a transmission time interval (TTI) basis. For each TTI a transport format is chosen and indicated through the transport format indicator (TFIN). In other words, the TFIN tells which transport format to use for that particular transport block on that particular TrCH during that particular TTI. When a transport channel is inactive, the transport format with a transport block size of zero (empty transport format) is selected.
Only a limited number of combinations of the transport formats of the different transport channels are allowed. A valid combination is called a Transport Format Combination (TFC). The set of valid TFCs on a basic physical subchannel is called a Transport Format Combination Set (TFCS). The TFCS is signalled through Calculated Transport Format Combinations (CTFC).
In order to decode a received sequence the receiver needs to know the active TFC for the radio packet. This information is transmitted in the Transport Format Combination Identifier (TFCI) field. Aforesaid field is basically a layer 1 header and has the same function as the stealing bits in GSM. Each of the TFC within a TFCS is assigned a unique TFCI value and upon receipt of a radio packet this is the first element to be decoded by the receiver. By exploiting the decoded TFCI value the transport formats for the different transport channels can be determined and the actual decoding can start.
In case of multislot operation, there shall be one FLO instance for each basic physical subchannel. Each FLO instance is configured independently by Layer 3 and gets an own TFCS as a result. The number of allocated basic physical subchannels depends on the multislot capabilities of the MS.
For the time being the use of FLO is planned to be limited to dedicated channels only, thus maintaining the 26-multiframe structure for which the SACCH shall be treated as a separate logical channel based on GERAN Release 5.
The concept of transport formats and channels as presented in reference [1] is visualized in FIG. 1 where e.g. coded speech is to be transmitted over FLO. Speech is transferred by using three different modes MODE 1, MODE 2, MODE 3 with different bit rates and an additional comfort noise generation mode CNG MODE. Inside a mode the speech bits have been divided into three different classes represented by three transport channels TrCHA 102, TrCHB 104, and TrCHC 106 on the basis of their varying importance during the speech reconstruction stage, for example. Numbers inside the blocks, see e.g. the block pointed by legend 108, being arbitrary in this example though, indicate the required number of bits in a transport channel and codec mode specific manner. Hence, it can be noticed from the figure that TrCHA contains four transport formats (0, 60, 40, 30), TrCHB three transport formats (0, 20, 40) and TrCHC only two formats (0, 20). Resulting transport format combinations TFC1-TFC4, that refer to transport formats on different channels that can be active at the same time, are depicted with dotted lines in the figure. All these valid combinations constitute the TFCS that is signalled through CTFC. An example of CTFC determination is found in reference [1] in addition to techniques applicable in proper TFC selection.
A protocol architecture of FLO in case of Iu mode is depicted in FIG. 2 wherein MAC layer 208 maps either a plurality of logical channels or TBFs (temporary block flows) from RLC entities located in RLC layer 206, said RLC layer 206 receiving data from e.g. PDCP 204 (Packet Data Convergence Protocol) and controlled by RRC (Radio Resource Controller) 202, to physical layer 210. In current specification [1] logical channels are used but are presumably to be replaced with the concept of temporary block flows in the future. The TBF concept is described in reference [3] in more detail. A dedicated channel (DCH) can be used as a transport channel dedicated to one MS in uplink or downlink direction. Three different DCHs have been introduced: CDCH (Control-plane DCH), UDCH (User-plane DCH) and ADCH (Associated DCH), the CDCH and UDCH of which used for transmission of RLC/MAC data transfer blocks, whereas the ADCH targeted for transmission of RLC/MAC control blocks. A mobile station may concurrently have a plurality of transport channels active.
The FLO architecture is illustrated in FIG. 3 especially in relation to Layer 1 for FLO. In this version only a one-step interleaving has been assumed, i.e. all transport channels on one basic physical subchannel have the same interleaving depth. An alternative architecture with two-step interleaving is disclosed in reference [1] for review. Basic error detection is carried out with a cyclic redundancy check. A Transport Block is inputted to error detection 302 that utilizes a selected generator polynomial in order to calculate the checksum to be attached to the block. Next, the updated block called Code Block is fed into a convolutional channel coder 304 introducing additional redundancy to it. In rate matching 306 bits of an Encoded Block are either repeated or punctured. As the block size can vary, also the number of bits on a transport channel may correspondingly fluctuate. Thereupon, bits shall be repeated or punctured in order to keep the overall bit rate in line with the actual allocated bit rate of the corresponding sub-channel. Output from rate matching block 306 is a called a Radio Frame. Transport channel multiplexing 308 takes care of multiplexing of Radio Frames from active transport channels TrCH(i) . . . TrCH(1) received from matching block 306 into a CCTrCH (Coded Composite Transport Channel). In TFCI mapping 310 a TFCI is constructed for the CCTrCH. Size of the TFCI depends on the number of TFCs needed. TFCI size should be minimized in order to avoid unnecessary overhead over the air interface. For example, TFCI of 5 bits can indicate 32 different transport format combinations. If these are not enough, a dynamic connection reconfiguration is needed to be performed. The TFCI is (block) coded and then interleaved 312 with CCTrCH (these two constituting a Radio Packet) on bursts. The selected interleaving technique is configured at call set-up. RRC layer, Layer 3 for FLO, manages set-up, reconfiguration and release of the traffic channels. Upon creating a new connection, Layer 3 indicates to the lower layers various parameters to configure the physical, MAC and RLC layers. Parameters include the transport channel identity (TrCH Id) and transport format set for each transport channel, transport format combination set through CTFC with modulation parameter etc. In addition, Layer 3 provides transport channel specific parameters such as CRC size, rate matching parameters, transport format dynamic attributes etc. The transport channels and the transport format combination set are separately configurable in the uplink and downlink directions.
Notwithstanding many advantageous features the FLO clearly offers to GERAN, some defects still remain in the proposed current version. For example, if a TFCS reconfiguration message is transmitted between a base station and MS but is lost or, at least, not correctly interpreted by the recipient, the subsequent communication may be endangered due to e.g. possible size difference between the previous and the new TFCI. Considering a scenario wherein the network informs the addition/deletion of TFCs in a TFCS reconfiguration message that is eventually lost, and the size of the TFCI is adapted as a result (addition->size increase, deletion->size decrease) at the sender side only, neither the network nor the MS is able to interpret the received data packets anymore due to the TFCIs with a different meaning to the two communication ends concerned. The similar problem arises if the far-end party properly receives the TFCS reconfiguration message but the indicated TFCS is taken into use at the ends of the communication without synchronization. Even if the TFCI size remains the same, different meanings between the old and the just received TFCIs forbids proper communication or at least disturbs it remarkably.